In the world of drug design, the key to activity is making a drug that will react on a cellular level at a specific site. One of the best ways to accomplish this is to synthesize a chiral ("handed") molecule that is site specific for a given receptor and has activity that blocks, inhibits or alters the normal chemistry at that site. In AIDS and cancer research for example, many active compounds that may prove valuable in eventually providing a cure have been found to be chiral. This chiral nature refers to the fact that the compound may exist as two asymmetric molecules that are mirror images of one another, i.e., they are related optically like right (D) and left (L) hands, hence the term "handed". However, since chemical reactions involving even pure (one isomer) chiral starting materials can yield mixtures of isomers (racemates), it is important to be able to isolate the desired chiral compound in its pure form since only one isomer will likely be active. This can be done in a variety of ways on a lab scale (crystallizations, chromatography, synthetic transformations, complexes), but most of these are not practical on a manufacturing scale.
Amino acids and peptides serve as important pharmaceutical and food additives in both human and animal diets and their production and purification have become vital to numerous food, drug and chemical industries. Many amino acids are made either chemically or through fermentation processes which require the separation and isolation of the desired amino acids from the broth.
Many amino acids exist as two optically active enantiomers, the L (levorotary) and D (dextrorotary) isomers. Most applications require the separation of the two from their racemic mixture that is produced during chemical manufacture. L-phenylalanine for example, is a precursor or component for many pharmaceutically active peptides and food additives and it exists in both the L and D forms. In many cases, the L form of the compound may be active whereas the D-isomer is not and visa versa. However, the synthesis of the amino acid or peptide in question produces a racemic mixture of both and hence it would be advantageous to find a way to isolate only the active isomer.
Various methods for separating the L and D isomers are known and generically are referred to as resolution. The most common method of resolving D,L-mixtures involves treating them with an optically active compound known as a resolving agent, followed by fractional crystallization of the resulting mixture of compounds (diastereroisomers) in solution. For practical resolution, it is necessary to find a combination of resolving agent and solvent which will give good crystallization behavior together with a pronounced difference in solubility between the diastereoisomers. Examples of this technique as applied to the resolution of an amino acid are described in U.S. Pat. No. 2,556,907 to Emmick and U.S. Pat. No. 2,657,230 to Rogers. These patents discuss methods to resolve D,L-lysine in which optically active glutamic acid is employed as the resolving agent.
More recently, phenylalanine has been resolved by the enzymatic hydrolysis of its racemic esters. The enzyme, chymotrypsin, selectively hydrolyzes L-phenylalanine esters. Hence, L-phenylalanine is recovered from a mixture of the D,L-phenylalanine ester. An example of this process is European Patent Appln. EP 174,862 to Empie Aug. 17, 1984.
The isoquinoline derivative (3S)-1,2,3,4-tetrahydroisoquinoline-3-carboxylic acid (L-TIC-A) is an amino acid that has shown to be a useful starting material in the enzymatic and chemical synthesis of a number of pharmaceutically active peptides and peptide analogues such as the bradykinin agonists, ACE inhibitors and hypotensive drugs. The D-form of the isomer is not as useful, however, as this isomer is relatively inactive. This presents a problem in that the synthesis of the compound from phenylalanine using the Pictet-Spengler reaction results in a mixture of both isomers. A method for the preparation of optically pure (3S)-1,2, 3,4-tetrahydroisoquinoline-3-carboxylic acid is therefore desirable.
In the Pictet-Spengler reaction,chiral (L) phenylalanine is reacted as set forth below to produce a major product, (L) TIC-A (85-90%) and a minor product, (D) TIC-A (15-10%). See Pictet, A; Spengler, T. Ber Dtsch Chem Ges 1911,44, 2030; and Archer, S. J Org. Chem 1951, 16, 430. The reaction occurs as follows. ##STR1##
This mixture of D,L TIC-A.cndot.HCl, must be neutralized and purified to isolate only (L) TIC-A, free acid, which is the "key" intermediate in the synthesis of the chiral product (L) N-t-Butyl-1,2,3,4-tetrahydro-3-isoquinoline carboxyamide (L-TIC-C) and ultimately (L) N-t-Butyl-decahydro-3-isoquinoline carboxamide (L-TIC-D) which are compounds useful as starting materials in the synthesis of a number of pharmaceutically active drugs. For example, N'-[1(S)benzyl-3-[4a(S),8a(S)-3-(S)-(tertbutylcarbamoyl)decahydroisoquinol ine-2-yl]-2(R)-hydroxypropyl]-N"-(quinolin-2-ylcarbonyl)-L-asparaginamide (Roche) is one derivative useful as an anti-viral and anti-HIV agent. Another compound, 3-Isoquinoline carboxylic acid, 2-[2-[(1-carboxy-3-phenylpropyl (amino]-1-oxopropyl]1,2,3,4-tetrahydro-[3-S[2[R(R),3R]-[CAS) (Parke Davis) is therapeutically useful as an antihypertensive. The purity of (L) TIC-A is crucial because none of the later intermediates can be practically purified at the manufacturing level if there is a racemic mixture consisting of the two L- and D-isomers.
Racemic and optically pure 1,2,3,4-tetrahydroisoquinoline-3-carboxylic acids and esters have also been prepared by the base-catalyzed cyclization reaction of 1,2-bis(halomethyl)benzenes with dimethyl 2-(acetylamino) malonate with subsequent decarboxylation and amide cleavage. Kammermaier et. al. Synthesis 1157 (November 1992) Enantiomer resolution is achieved by esterification with menthol followed by column chromatography or by distereomeric salt separation of the benzylic ester with mandelic acid and a base catalyzed saponification.
Shiraiwa et. al., Bull. Chem. Soc. Jpn. 64 3729 (December 1991) reports the asymmetric transformation of L-TIC-A through the use of (1S)-10-camphor-sulfonic acid as a resolving agent that yields a salt of L-TIC-A with 90% optical purity. The TIC obtained from the salt was purified to give a yield of optically pure L-TIC-A of about 80%.
EPA 496 369 to Kammermaier discloses a method for the preparation of racemic TIC-A whereby dihalo-o-xylene is cyclized with N-acylamidomalonic acid dialkyl esters in a basic medium to dicarboxylic esters which are subsequently saponified and acidified to the final product.
EPA 049 658 to Remond et. al. discloses the preparation of (3S)-tetrahydroisoquinoline-3-carboxylic acid through the treatment of (S).beta.-phenylalanine with concentrated hydrochloric acid and a 40% formaldehyde solution. After heating and subsequently cooling, the compound in its racemic forms crystallizes out.
U.S. Pat. No. 4,847,409 to Kidman et. al. discloses a method for the isolation of a substantially pure L-isomer of an amino acid from its D,L-mixture that does not require the use of a resolving agent, the formation of a derivative of the amino acid or additional enzymatic reactions. The method is based on the equal saturation point whereby the L-isomer can be precipitated out and separated from the D-isomer when the relative concentration of the D-isomer is less than 7.0%.
As mentioned earlier, several basic ways of purifying these and other isomers are known: crystallization (often several are necessary to give a pure isomer); chromatography; synthetic transformation whereby the compound mixture is first converted into one that is more easily purified, it is purified and the purified material is returned to its original form; or complexation of the desired isomer is carried out with a resolving reagent that yields a more easily separated mixture, with subsequent removal of the resolving agent. All of these methods will work, but at a price, either higher synthetic costs and/or lower overall yields. Moreover, few if any at all are feasible for a large scale manufacturing process.
Crystallization of the racemic mixture is relatively straight forward on the laboratory scale (see the Journal of Medicinal Chemistry, 1988, Vol. 31, No. 11, 2095-2096), but several crystallizations are usually required and this method is not feasible on a manufacturing scale. Synthetic transformation from the "acid" form to the "ester salt" form is another purification technique (Chem. Pharm. Bull., 31,313-14, 1983) which requires synthesis of the ester salt followed by fractional crystallization and subsequent de-esterification. This invariably results in extra expense in time/labor/materials necessary to obtain a pure product. Asymmetric transformation of racemic mixtures by use of a resolving agent (Bull. Chem Soc. Jpn. 64, 3729-31, 1991) requires a specific resolving agent which must be reacted with the mixture and later removed from the pure isomer, which again, results in extra expense. Chromatography is an excellent separation method that is an essential small, lab-scale tool but which seldom can be utilized for purification of isomers on a large manufacturing scale.
In all of these conventional prior art methods, additional steps involving the use of a resolving agent in a chemical or enzymatic reaction or some combination of these is required to form derivatives of L-TIC-A in order to isolate the desired isomer. The present invention permits the separation of an optically pure L-isomer from the D,L-mixture and in the same step, purifies it from other amino acids, salts, etc. which are manufacturing contaminants, without the requirement of a resolving agent, without the subsequent formation of a derivative of the amino acid in question and without enzymatic reaction. There is also no need as is found in the methods known in the art of first converting the free acid to the ester, purifying it and then reconverting the ester to the optically pure free acid. These additional steps and intermediates are thereby done away with resulting in a more economically efficient, cleaner product.